Consumer Sleep Technologies Research Articles

Public engagement with consumer sleep technology for obstructive sleep apnea screening: implications for equity, access, and practice.

To characterize public practices and perspectives on the use of consumer sleep technology (CST) and evaluate perspectives on using CST as a screening tool for obstructive sleep apnea (OSA). We designed a survey instrument incorporating content from validated instruments (STOP-BANG and the Epworth Sleepiness Scale) and hypothesis-generated questions. Survey development involved multidisciplinary collaboration among three board-certified sleep medicine experts, researchers, and consumers. The survey was disseminated across a national sample of adults living in the United States via an online platform. Among 897 respondents, the mean (SD) age was 47.5 (16.9) years; 73.1% were female, 81.8% were White, and 505 respondents (56.3%) reported having tracked sleep using CS. Factors associated with decreased odds of CST use included household income <$30,000 (OR 0.47, 95% CI 0.28-0.79; p=0.004), Medicaid insurance (OR 0.43, 95% CI 0.26-0.69; p=0.001), Medicare insurance (OR 0.59, 95% CI 0.41-0.84; p=0.004), and lack of a primary care physician (OR 0.55, 95% CI 0.33-0.91; p=0.021). Most respondents (91.1%) agreed or strongly agreed that screening for OSA would be a useful feature of CST, but respondents reporting an education of high school diploma or less (OR 0.48, 95% CI 0.29-0.79; p=0.004) were less likely to agree with this statement. Attitudes toward and use of CST differed based on demographic and socioeconomic factors. Further study is needed to understand and address barriers to CST adoption and to characterize implications for equitable access to care for sleep disorders.

Sleep Behavior in Royal Australian Navy Shift Workers by Shift and Exposure to the SleepTank App.

Rotating shiftwork schedules are known to disrupt sleep in a manner that can negatively impact safety. Consumer sleep technologies (CSTs) may be a useful tool for sleep tracking, but the standard feedback provided by CSTs may not be salient to shift-working populations. SleepTank is an app that uses the total sleep time data scored by a CST to compute a percentage that equates hours of sleep to the fuel in a car and warns the user to sleep when the "tank" is low. Royal Australian Navy aircraft maintenance workers operating on a novel rotational shift schedule were given Fitbit Versa 2s to assess sleep timing, duration, and efficiency across a 10-week period. Half of the participants had access to just the Fitbit app while the other half had access to Fitbit and the SleepTank app. The goal of this study was to evaluate differences in sleep behavior between shifts using an off-the-shelf CST and to investigate the potential of the SleepTank app to increase sleep duration during the 10-week rotational shift work schedule. Royal Australian Navy volunteers agreed to wear a Fitbit Versa 2 with the SleepTank app (SleepTank condition), or without the SleepTank app (Controls), for up to 10 weeks from May to July 2023 during the trial of a novel shift rotation schedule. Participants from across 6 units worked a combination of early (6:00 AM to 2:00 PM), day (7:30 AM to 4:30 PM), late (4:00 PM to 12:00 AM), and night shifts (12:00 AM to 6:00 AM) or stable day shifts (6:00 AM to 4:00 PM). Differences in sleep behavior (time in bed, total sleep time, bedtime, wake time, sleep efficiency [SE]) between conditions and shift types were tested using Analysis of Variance. This study was approved by the Australian Departments of Defence and Veterans' Affairs Human Research Ethics Committee. Thirty-four participants completed the full study (n = 17 Controls; n = 17 SleepTank). There was a significant effect of shift type on 24-hour time in bed (TIB24; F(4,9) = 8.15, P < .001, η2 = 0.15) and total sleep time (TST24; F(4,9) = 8.54, P < .001, η2 = 0.18); both were shorter in early shifts and night shifts compared to other shift types. TIB24 and TST24 were not significantly different between conditions, but there was a trend for greater SE in the SleepTank condition relative to Controls (F(1,9) = 2.99, P = .08, η2 = 0.11). Sleep data collected by Fitbit Versa 2s indicated shorter sleep duration (TIB24, TST24) for Royal Australian Navy workers during early and late shifts relative to stable day shifts. Access to the SleepTank app did not greatly influence measures of sleep duration but may be protective against fatigue by affecting SE. Further research is needed to evaluate the utility of the SleepTank app as a means of improving sleep hygiene in real-world, shift-working environments.

A Deep Transfer Learning Approach for Sleep Stage Classification and Sleep Apnea Detection Using Wrist-Worn Consumer Sleep Technologies.

Obstructive sleep apnea (OSA) is a common, underdiagnosed sleep-related breathing disorder with serious health implications Objective - We propose a deep transfer learning approach for sleep stage classification and sleep apnea (SA) detection using wrist-worn consumer sleep technologies (CST). Methods - Our model is based on a deep convolutional neural network (DNN) utilizing accelerometers and photo-plethysmography signals from nocturnal recordings. The DNN was trained and tested on internal datasets that include raw data from clinical and wrist-worn devices; external validation was performed on a hold-out test dataset containing raw data from a wrist-worn CST. Results - Training on clinical data improves performance significantly, and feature enrichment through a sleep stage stream gives only minor improvements. Raw data input outperforms feature-based input in CST datasets. The system generalizes well but performs slightly worse on wearable device data compared to clinical data. However, it excels in detecting events during REM sleep and is associated with arousal and oxygen desaturation. We found; cases that were significantly underestimated were characterized by fewer of such event associations. Conclusion - This study showcases the potential of using CSTs as alternate screening solution for undiagnosed cases of OSA. Significance - This work is significant for its development of a deep transfer learning approach using wrist-worn consumer sleep technologies, offering comprehensive validation for data utilization, and learning techniques, ultimately improving sleep apnea detection across diverse devices.

Fatigue Risk Management Preferences for Consumer Sleep Technologies and Data Sharing in Aviation.

INTRODUCTION: Employees from any type of aviation services industry were asked to give their opinions about the usefulness of consumer sleep technologies (CSTs) during operations and their willingness to share data from CSTs with their organizations for fatigue risk management purposes under a variety of circumstances.METHODS: Respondents provided information about position in aviation and use of CST devices. Respondents ranked sleep issues and feedback metrics by perceived level of importance to operational performance. Respondents rated their likelihood to share data with their organization under a series of hypothetical situations.RESULTS: Between January-July 2023, 149 (N = 149) aviation professionals responded. Pilots comprised 72% (N = 108) of respondents; 84% (N = 125) of all respondents worked short- or medium-haul operations. "Nighttime operations" and "inconsistent sleep routines" ranked as the most important issues affecting sleep. "Sleep quality history" and "projected alertness levels" ranked as most important feedback metrics for personal management of fatigue. Respondents were split between CST users (N = 64) and nonusers (N = 68). CST users did not indicate a strong preference for a specific device brand. The most-reported reason for not using a CST was due to not owning one or no perceived need. Respondents indicated greater likelihood of data sharing under conditions where the device was provided to them by their organization.DISCUSSION: These results suggest that aviation professionals are more concerned about schedule-related disturbances to sleep than they are about endogenous sleep problems. Organizations may be able to increase compliance to data collection for fatigue risk management by providing employees with company-owned CSTs of any brand.Devine JK, Choynowski J, Hursh SR. Fatigue risk management preferences for consumer sleep technologies and data sharing in aviation. Aerosp Med Hum Perform. 2024; 95(5):265-272.

Impact of Hypoglossal Nerve Stimulation on Consumer Sleep Technology Metrics and Patient Symptoms.

Obstructive sleep apnea (OSA) is usually assessed at discrete and infrequent timepoints. Wearable consumer sleep technologies (CST) may allow for more granular and longitudinal assessments of OSA therapy responses and OSA-related symptoms. In this case series, we enrolled hypoglossal nerve stimulator (HGNS) patients who had an effective treatment response for an 8-week study using a wearable CST. Participants started with "HGNS-on," were randomized to turn off HGNS therapy during either week 4 or 5 ("HGNS-off"), followed by a return to therapy, "HGNS-resume." Participants completed validated symptom questionnaires assessing sleepiness, insomnia symptoms, functional status, and overall sleep health (Satisfaction, Alertness, Timing, Efficiency, and Duration, SATED) each week. CST metrics and survey scores were compared between HGNS treatment phases. Associations between CST metrics and survey scores were assessed. Seven participants with a total of 304 nights of CST data showed no statistically significant changes in total sleep time (TST), wake time after sleep onset, or sleep efficiency (SE) across the study periods. During HGNS-off, survey scores indicated significantly worsened OSA-related symptom scores. Two participants had significantly higher heart rate variability (HRV) during HGNS-off (by 3.3 and 6.3 ms) when compared to HGNS active therapy periods. Amongst CST metrics, SATED scores correlated with TST (r = 0.434, p < 0.0001), HRV (r = -0.486, p < 0.0001), and SE (r = 0.320, = 0.0014). In addition, FOSQ-10 scores correlated with average HR during sleep (r = -0.489, p < 0.001). A 1-week HGNS therapy withdrawal period impacted OSA-related sleep symptoms. Sleep-related metrics measured by a wearable CST correlated with symptom scores indicating potential value in the use of CSTs for longitudinal sleep-tracking in OSA patients. 4 Laryngoscope, 134:3406-3411, 2024.

Guidelines for the development, performance evaluation and validation of new sleep technologies (DEVSleepTech guidelines) - a protocol for a Delphi consensus study.

New sleep technologies are being developed, refined and delivered at a fast pace. However, there are serious concerns about the validation and accuracy of new sleep-related technologies being made available, as many of them, especially consumer-sleep technologies, have not been tested in comparison with gold-standard methods or have been approved by health regulatory agencies. The importance of proper validation and performance evaluation of new sleep technologies has already been discussed in previous studies and some recommendations have already been published, but most of them do not employ standardized methodology and are not able to cover all aspects of new sleep technologies. The current protocol describes the methods of a Delphi consensus study to create guidelines for the development, performance evaluation and validation of new sleep devices and technologies. The resulting recommendations are not intended to be used as a quality assessment tool to evaluate individual articles, but rather to evaluate the overall procedures, studies and experiments performed to develop, evaluate performance and validate new technologies. We hope these guidelines can be helpful for researchers who work with new sleep technologies on the appraisal of their reliability and validation, for companies who are working on the development and refinement of new sleep technologies, and by regulatory agencies to evaluate new technologies that are looking for registration, approval or inclusion on health systems.

Consumer preferences for sleep-tracking wearables: The role of scientific evaluation and endorsement

ObjectivesAccuracy and relevance to health outcomes are important to researchers and clinicians who use consumer sleep technologies, but economic demand motivates consumer sleep technology design. This report quantifies the value of scientific relevance to the general consumer in a dollar amount to convey the importance of device accuracy in terms that consumer sleep technology manufacturers can appreciate. MethodsSurvey data were collected from 368 participants on Amazon mTurk. Participants ranked sleep metrics, evaluation methods, and scientific endorsement by perceived level of importance. Participants indicated their likelihood of purchasing a hypothetical consumer sleep technology that had either (1) not been evaluated or endorsed; (2) had been evaluated but not endorsed, and; (3) had been evaluated and endorsed by a sleep science authority. Demand curves determined the relative value of each consumer sleep technology. ResultsDevices that were evaluated and endorsed had the most value, followed by those only evaluated, and then those with no evaluation. The unit price at which there was 50% probability of purchase increased by $30 or $48 for evaluation or endorsement, respectively, relative to a nonvalidated device. Respondents indicated the most valuable sleep metric was sleep duration, the most important evaluation method was against laboratory/hospital standards for sleep, and that the highest value of endorsement came from a medical institution. ConclusionsConsumer demand is greatest for a device that has been evaluated by an independent laboratory and is endorsed by a medical institution. Consumer sleep technology manufacturers may be able to increase sales by partnering with sleep science authorities to produce a scientifically superior device.

Clinical Evaluation of the Sleepy and Sleepless Patient.

This article addresses the approach to the evaluation of patients who present to a neurologist with excessive daytime sleepiness or difficulty sleeping. Greater emphasis on the importance of sleep reflects the growing scientific understanding that sleep is critical to overall health and well-being. Consumer sleep technologies, which measure parameters related to sleep, may provide insight into an individual's sleep-related symptoms and tendencies and have a role in patient-centered sleep evaluation when used within an appropriate clinical context. A thorough review of a patient's history and physical examination findings are important components of the assessment and management of their sleep-related symptoms. An understanding of how the clinical context relates to the categorization of sleep disorders can impact a patient's symptoms, comorbid neurologic disorders, and overall well-being. Many neurologic conditions are strongly associated with sleep disturbance, risk factors for the development of a sleep disorder, or both. Therefore, it is critical for neurologists to be familiar and comfortable with taking a focused sleep history. Modalities such as in-laboratory polysomnography, home sleep apnea testing, multiple sleep latency testing, and actigraphy, as well as contextualized and prudent use of data obtained from consumer sleep technologies, can be helpful in appropriately selected patients. Mindful integration of these objective data facilitates the diagnosis and management of sleep disorders.

0279 Sleep Hygiene for Sleep Health in the General Population: What Does Data From Consumer Sleep Technology Tell Us?

Abstract Introduction Despite being used and widely recommended since the 1970s, few studies have examined whether adherence to sleep hygiene practices affect objectively measured sleep in non-clinical populations. While individual components of sleep hygiene such as limiting caffeine and alcohol consumption are clearly related to sleep by plausible physiological and psychosocial mechanisms, the real-world evidence of overall sleep hygiene practices on sleep is surprisingly inconsistent. Here, we examined the association between self-reported sleep hygiene practices and objectively measured sleep in a general population. Methods Responses to a survey on sleep hygiene were used and matched with objective sleep data, resulting in data from 720 users (mean age: 52.5 ± 15.9, 63.4% female). Objective sleep data across 92,808 nights were included in the analysis from the PSG-validated SleepScore Mobile Application, which uses a non-contact sonar-based method to capture sleep-related metrics and self-reported lifestyle. Self-reported sleep hygiene practices were assessed with 13-items on a 5-point scale ranging from “Never” to “Always”. Descriptive statistics and linear regressions were used for the analysis, controlling for age and gender. Results Overall, the top three most frequented poor sleep hygiene practices were going to bed at different times (29.7%), overthinking/worrying in bed (24.0%), and waking at different times (22.7%). Linear regressions revealed a significant negative association between composite sleep hygiene scores and objective sleep measures, whereby poorer sleep hygiene was associated with significant reductions in total sleep time (ß=-0.89, SE=0.33, p&amp;lt; 0.01), REM duration (ß=-0.24, SE=0.10, p&amp;lt; 0.05), and SleepScore (ß=-0.15, SE=0.05, p&amp;lt; 0.01), an objective sleep quality metric. No significant associations were observed between individual sleep hygiene factors and objectively measured sleep. Conclusion While we could not identify a relationship between individual hygiene factors and sleep, poorer aggregated sleep hygiene scores were associated with poorer objectively measured sleep. Thus, sleep health may not be defined by one single behavior, but rather by the sum of its parts. Future work should examine the efficacy of personalized sleep hygiene factors in sub-clinical populations, where targeted sleep hygiene education may be preferred given it is more intuitive and less burdensome than other behavioral interventions. Support (if any) SleepScore Labs

0375 Increased Engagement with a Tailored Sleep Guide is Associated with Reduced Variability in Sleep Efficiency in Chronic Insomnia

Abstract Introduction Insomnia is highly prevalent, but difficult to diagnose due to night-to-night sleep variability, and difficult to treat in part due to the lack of trained providers. Consumer sleep technology (CST) allows for longitudinal sleep monitoring in a natural environment. One such device, the Sleep Score Max (SleepScore Labs) combines longitudinal non-contact sleep tracking via radio-frequency biomotion sensor technology with an individualized sleep guide function, which provides tailored recommendations to improve sleep based on the user’s objectively measured sleep, bedroom environment, and self-reported daily activity. Focusing on sleep efficiency, we examined if level of engagement with the sleep guide is related to sleep improvement in chronic insomnia. Methods As part of a larger study, N=30 individuals with chronic insomnia (ages 38.7±14.9; 8 male) participated in an 8-week at-home sleep monitoring study. Enrollment criteria included being 18-65y old and meeting ICSD-3 criteria for chronic insomnia with no other clinically relevant condition contributing to sleep disturbance. Participants were instructed to track their sleep with the aforementioned device and engage with the sleep guide daily for 8 consecutive weeks. For observed sleep efficiency, means (characterizing sleep overall) and within-subject standard deviations (characterizing ­night-to-night variability) were assessed by week, and change scores from week 1 to week 8 were correlated with level of engagement with the sleep guide function (defined as percentage of days interacted with the guide within the 8-week period). Results From week 1 to week 8, mean sleep efficiency increased non-significantly (t[23]=0.83, p=0.42), but night-to-night variability decreased significantly (t[23]=-2.16, p=0.041). There was no significant relationship between change in mean sleep efficiency and level of engagement with the sleep guide (r=0.005, p=0.980), but there was a significant relationship between change in night-to-night variability and level of engagement (r=-0.48, p=0.018), such that night-to-night ­variability decreased more in those with greater engagement. Conclusion Night-to-night sleep variability is a chief complaint of people with chronic insomnia, and reduced night-to-night variability in sleep efficiency may indicate improvement. As such, engagement with an individualized sleep guide incorporated in a CST device may help individuals with chronic insomnia improve their sleep. Support (if any) NIH grant KL2TR002317; research devices provided by SleepScore Labs.

0299 Sleep Hygiene in Nurses Using Data Collected from Consumer Sleep Technology (CST) with the SleepTank™ mobile app

Abstract Introduction Nurses are prone to fatigue and sleep disturbances due to extended hour or shiftwork schedules. Fatigue in nurses not only compromises their own health and safety, but jeopardizes patient care as well. Nurses' work hours and overtime are currently not federally regulated. Fatigue risk mitigation strategies for nurses would benefit from understanding the relationship between sleep duration and work patterns in order to develop tools and regulations to protect against fatigue in the nursing profession. Methods Thirteen (N=13) nurses wore a Fitbit Versa 2 and used the SleepTank™ mobile app during a 6-week work rotation. Nurses in this study selected their own work shifts but did have duty-hour limitations. Sleep data was extracted via the mobile app; work schedules were self-reported. Sleep duration data was summed by date to provide a 24-hour estimate of time in bed (TIB24). Repeated measures analysis of variance (rmANOVA) explored differences in TIB24 by number of consecutive days worked and time. Pearson’s correlations evaluated the relationship between work and sleep. Results Nurses (N=13) wore the device with the app for an average of 54±15 days. Forty percent (40%) of the study period were work days (22±10). Nurses slept longer on non-work days (525±153 min) compared to work days (481±143; t=3.21, p=0.001). Nurses worked 2 days in a row on average (Range: 1-5). TIB24 was ~30 minutes shorter with each consecutive day worked (F5, 81=8.11, η2=0.08, p=&amp;lt; 0.001). Routinely working a greater number of consecutive days was negatively correlated to TIB24 (r=0.71, p=0.009). Later average work start time was positively correlated with later bedtime (r=0.83, p&amp;lt; 0.001) and shorter TIB24 (r=0.65, p=0.01). Conclusion Working more days consecutively and working later shifts was related to shorter sleep duration and later bedtime in nurses. However, given that nurses in this study routinely slept longer than 7 hours per night, poor sleep hygiene does not appear to be an issue in this population. The ability to select their own work shifts may allow nurses to select a schedule that permits healthy sleep behavior. Support (if any) N/A

0952 A Long Winter’s Sleep: What Does Big Data From Consumer Sleep Technology Tell Us?

Abstract Introduction Light is the most powerful zeitgeber influencing sleep-wake patterns, and winter’s shorter days have profound effects on mood, energy, and metabolic function. Here, we leverage big data from consumer sleep technology to examine the relationship between objectively measured sleep in winter versus summer months in the United States. Methods The dataset included 21,101 users residing in the US across 1,255,518 nights (age range: 17-90, mean age: 46.6 ± 16.7 years, 59% female). Data were extracted from 01 January 2019 until 31 August 2022. Seasons were defined according to the Northern Hemisphere’s meteorological calendar and included winter (December - February) and summer (June - August). USA public holidays were mapped and removed from the data. The relationships between objectively measured sleep and winter versus summer were examined using linear regression models, controlling for age and gender. Results A shift to later average monthly wake times were observed in both summer and winter, peaking in summer. Similarly, bedtimes shift later in summer and winter, yet the shift in summer was greater than winter (winter vs summer: waketime: 07:06 vs 07:07, ß = -0.46, SE = 0.023, p&amp;lt; 0.001, bedtime: 23:17 vs 23:23, ß = 0.12, SE = 0.006, p&amp;lt; 0.001). Although bedtime and wake times delay during the winter and summer seasons, measures of sleep health including sleep duration, sleep efficiency, and SleepScore peak in winter and slowly decline to their lowest point during the summer months (winter vs. summer: SleepScore 78.4 vs 77.6, ß = -0.89, SE = 0.17, p&amp;lt; 0.001; sleep duration 359.6 min vs 352.8 min, ß = -4.8, SE = 1.02, p&amp;lt; 0.001; sleep efficiency 77.5 vs 76.6, ß = 0.86, SE = 0.038, p&amp;lt; 0.001). Conclusion Bedtime and wake time delays were greater in summer than winter, likely contributing to small improvements to sleep quality, duration, and efficiency in winter compared to summer. These findings are generally supported by previous research examining sleep-wake patterns in locations with extreme seasonal differences, such as Northern Norway, where, despite midnight sun and polar nights, researchers have found little variation in sleep-wake timing. Support (if any) SleepScore Labs

Smart, Personalized Sleep: The Benefits and Risks of Artificial Intelligence and Wearables in the Consumer Sleep Technology Market.

Smart, Personalized Sleep: The Benefits and Risks of Artificial Intelligence and Wearables in the Consumer Sleep Technology Market.

Sleep medicine provider perceptions and attitudes regarding consumer sleep technology.

This study assessed perceptions and attitudes of sleep medicine providers regarding consumer sleep technology (CST). A convenience sample of n = 176 practicing sleep medicine and behavioral sleep medicine experts was obtained using social media and the American Academy of Sleep Medicine directory. Providers completed a questionnaire that assessed perceptions and attitudes about patient use of CST in the clinical setting. The sample included both adult and pediatric psychologists, physicians, and advanced practice providers from a variety of health settings. Providers reported 36% (3%-95%) of patients used CST, and the most common devices seen by providers were wrist-worn devices followed by smartphone apps. The most common perceived patient motivations for frequent use were to measure sleep and self-discovery. Across sleep disorders, clinicians did not endorse frequent CST use; the highest reported use was for assisting patients in the completion of sleep diaries. Overall devices were rated as somewhat accurate and neutral regarding helpfulness. In qualitative responses, providers associated CST use with increased patient engagement but increased orthosomnia and misperceptions about sleep. CST is frequently encountered in the sleep medicine clinic, and providers view CST as somewhat accurate but neither helpful nor unhelpful in clinical practice. Although providers viewed these devices as useful to drive patient engagement/awareness and track sleep patterns, providers also viewed them as a contributor to orthosomnia and misperceptions about sleep. Addison C, Grandner MA, Baron KG. Sleep medicine provider perceptions and attitudes regarding consumer sleep technology. J Clin Sleep Med. 2023;19(8):1457-1463.

The good, the bad, and the ugly of consumer sleep technologies use among athletes: A call for action

The good, the bad, and the ugly of consumer sleep technologies use among athletes: A call for action

Consumer sleep technology for the screening of obstructive sleep apnea and snoring: current status and a protocol for a systematic review and meta-analysis of diagnostic test accuracy.

There are concerns about the validation and accuracy of currently available consumer sleep technology for sleep-disordered breathing. The present report provides a background review of existing consumer sleep technologies and discloses the methods and procedures for a systematic review and meta-analysis of diagnostic test accuracy of these devices and apps for the detection of obstructive sleep apnea and snoring in comparison with polysomnography. The search will be performed in four databases (PubMed, Scopus, Web of Science, and the Cochrane Library). Studies will be selected in two steps, first by an analysis of abstracts followed by full-text analysis, and two independent reviewers will perform both phases. Primary outcomes include apnea-hypopnea index, respiratory disturbance index, respiratory event index, oxygen desaturation index, and snoring duration for both index and reference tests, as well as the number of true positives, false positives, true negatives, and false negatives for each threshold, as well as for epoch-by-epoch and event-by-event results, which will be considered for the calculation of surrogate measures (including sensitivity, specificity, and accuracy). Diagnostic test accuracy meta-analyses will be performed using the Chu and Cole bivariate binomial model. Mean difference meta-analysis will be performed for continuous outcomes using the DerSimonian and Laird random-effects model. Analyses will be performed independently for each outcome. Subgroup and sensitivity analyses will evaluate the effects of the types (wearables, nearables, bed sensors, smartphone applications), technologies (e.g., oximeter, microphone, arterial tonometry, accelerometer), the role of manufacturers, and the representativeness of the samples.

A Flexible Deep Learning Architecture for Temporal Sleep Stage Classification Using Accelerometry and Photoplethysmography.

Wrist-worn consumer sleep technologies (CST) that contain accelerometers (ACC) and photoplethysmography (PPG) are increasingly common and hold great potential to function as out-of-clinic (OOC) sleep monitoring systems. However, very few validation studies exist because raw data from CSTs are rarely made accessible for external use. We present a deep neural network (DNN) with a strong temporal core, inspired by U-Net, that can process multivariate time series inputs with different dimensionality to predict sleep stages (wake, light-, deep-, and REM sleep) using ACC and PPG signals from nocturnal recordings. The DNN was trained and tested on 3 internal datasets, comprising raw data both from clinical and wrist-worn devices from 301 recordings (PSG-PPG: 266, Wrist-worn PPG: 35). External validation was performed on a hold-out test dataset containing 35 recordings comprising only raw data from a wrist-worn CST. An accuracy = 0.71 ± 0.09, 0.76 ± 0.07, 0.73 ± 0.06, and κ = 0.58 ± 0.13, 0.64 ± 0.09, 0.59 ± 0.09 was achieved on the internal test sets. Our experiments show that spectral preprocessing yields superior performance when compared to surrogate-, feature-, raw data-based preparation. Combining both modalities produce the overall best performance, although PPG proved to be the most impactful and was the only modality capable of detecting REM sleep well. Including ACC improved model precision to wake and sleep metric estimation. Increasing input segment size improved performance consistently; the best performance was achieved using 1024 epochs (∼8.5 hrs.). An accuracy = 0.69 ± 0.13 and κ = 0.58 ± 0.18 was achieved on the hold-out test dataset, proving the generalizability and robustness of our approach to raw data collected with a wrist-worn CST.

Agreement of Sleep Measures-A Comparison between a Sleep Diary and Three Consumer Wearable Devices.

Nowadays, self-tracking and optimization are widely spread. As sleep is essential for well-being, health, and peak performance, the number of available consumer technologies to assess individual sleep behavior is increasing rapidly. However, little is known about the consumer wearables’ usability and reliability for sleep tracking. Therefore, the aim of the present study was to compare the sleep measures of wearable devices with a standardized sleep diary in young healthy adults in free-living conditions. We tracked night sleep from 30 participants (19 females, 11 males; 24.3 ± 4.2 years old). Each wore three wearables and simultaneously assessed individual sleep patterns for four consecutive nights. Wearables and diaries correlated substantially regarding time in bed (Range CCCLin: 0.74–0.84) and total sleep time (Range CCCLin: 0.76–0.85). There was no sufficient agreement regarding the measures of sleep efficiency (Range CCCLin: 0.05–0.34) and sleep interruptions (Range CCCLin: −0.02–0.10). Finally, these results show wearables to be an easy-to-handle, time- and cost-efficient alternative to tracking sleep in healthy populations. Future research should develop and empirically test the usability of such consumer sleep technologies.

Expert Demand for Consumer Sleep Technology Features and Wearable Devices: A Case Study

Global demand for sleep-tracking wearables, or consumer sleep technologies (CSTs), is steadily increasing. CST marketing campaigns often advertise the scientific merit of devices, but these claims may not align with consensus opinion from sleep research experts. Consensus opinion about CST features has not previously been established in a cohort of sleep researchers. This case study reports the results of the first survey of experts in real-world sleep research and a hypothetical purchase task (HPT) to establish economic valuation for devices with different features by price. Forty-six (N = 46) respondents with an average of 10 ± 6 years’ experience conducting research in real-world settings completed the online survey. Total sleep time was ranked as the most important measure of sleep, followed by objective sleep quality, while sleep architecture/depth and diagnostic information were ranked as least important. A total of 52% of experts preferred wrist-worn devices that could reliably determine sleep episodes as short as 20 min. The economic value was greater for hypothetical devices with a longer battery life. These data set a precedent for determining how scientific merit impacts the potential market value of a CST. This is the first known attempt to establish a consensus opinion or an economic valuation for scientifically desirable CST features and metrics using expert elicitation.

0607 Prevalence and Correlates of Sleep Disorders Among Users of a Consumer Sleep Technology

Abstract Introduction Sleep disorders constitute a major public health burden yet remain widely undiagnosed and untreated. The prevalence of diagnosed sleep disorders and associations with objectively measured sleep-wake dysfunction vary widely across populations. Here, we examined the self-reported prevalence and objective sleep architectural correlates of four sleep disorders in a large sample of individuals using a consumer sleep technology. Methods Data from 33,429 users (mean age: 44.6, 55.1% female) across 1,842,282 nights were included in the analysis from the PSG-validated SleepScore Mobile Application, which uses a non-contact, sonar-based method to objectively capture sleep-related metrics, and questionnaires to capture self-reported data. Subjective sleep disorder information was ascertained by asking users, “Which of the following sleep disorders has a healthcare professional diagnosed you with?” Linear regression was used for analysis, while age and gender were used as confounding variables, with the cohort reporting “None of the above” were used reference group for research purposes. Results The prevalence of reported disorders were “None of the above” (n=23,732,71.0%), sleep apnea/SDB (n=5,309, 15.9%), insomnia (n=3,968, 11.9%), RLS/PLM (n=2,295, 6.87%), or narcolepsy (n=266, 0.8%). Narcolepsy was associated with the greatest reduction in TST (ß=-23.6 mins, SE=3.475, p&amp;lt;0.001) while insomnia was associated with smallest (ß=-5.7mins, SE=0.979, p&amp;lt;0.001). Narcolepsy was associated with the greatest increase in WASO (ß=7.0 mins, SE=1.815, p&amp;lt;0.001) while insomnia was associated with smallest (ß=2.2 mins, SE=0.511, p&amp;lt;0.001). RLS/PLM was associated with the greatest increase in SOL (ß=3.9mins, SE=0.302, p&amp;lt;0.001) while sleep apnea/SDB was associated with the smallest (ß=2.171 mins, SE=0.22, p&amp;lt;0.001). Narcolepsy was associated with the greatest decrease in SE (ß=-3.05%, SE=0.5, p&amp;lt;0.001) while insomnia was associated with smallest (ß=-1.42%, SE=0.1, p&amp;lt;0.001). Conclusion Self-reported sleep disorders were associated with objectively poor sleep in a big data consumer sleep technology analysis. These findings suggest consumer sleep technologies may have value in screening for sleep disorders in the general population and may motivate these individuals to seek care in clinical sleep medicine settings. Support (If Any)